The task allocation for multiple UAVs is a complex multi-constraint,multi-objective,nonlinear optimization problem.Aiming at the defect of each commonly-used intelligent algorithm,and based on the characteristics of fireworks algorithm,this paper proposes an improved adaptive fireworks algorithm to deal with the problem.The mathematical model of the problem is built firstly,then,the weight of each index is obtained by using AHP together with the gray correlation method.The model is then transformed into a problem of unconstrained,single-objective extremum value by using the external penalty function method.After that,the adaptive fireworks algorithm is used to solve the model.In order to verify the superiority of adaptive fireworks algorithm in dealing with this problem,several different optimization algorithms are used for simulation.The experimental simulation results show that,the adaptive fireworks algorithm can quickly converge to the global optimum solution,and the result intuitively describes the reasonable task allocation scheme under complex situations.